International Choice Modelling Conference, International Choice Modelling Conference 2017

Font Size: 
Map matters: using individually tailored maps to value preferences for agricultural land use change in Great Britain
Tomas Badura, Silvia Ferrini, Ian Bateman, Amy Binner, Michael Burton

Last modified: 28 March 2017


Existing literature in environmental valuation have struggled to incorporate the spatial context of values, often omitting what is in fact one of their key determinants. This study develops a novel approach to addressing this problem while looking at the values people associate with changes in the agricultural landscape that have direct impact on biodiversity. The novelty of the paper lies in its choice set presentation, which follows a spatially tailored web-survey design with over 2,500 respondents and offers means for assessing the consistency, comprehension and reliability of CE responses and results.

Academics, policy-makers and NGO sectors alike have voiced the need to take account of the importance of natural capital, and the full value of benefits nature provides to people, in decision making and accounting systems.  In fact, the lack of appreciation of the myriad ways nature contributes to human well-being represents one of the drivers of its loss. This is most obvious in land use policies where decisions omitting natural capital can lead to sub optimal use of land to the detriment to both society and the environment. As a result, understanding the values associated with land use change and its impacts on people’s welfare could help to develop better policies. In this study, we explored land use change in Great Britain, where around 70% of land is agricultural and where loss of biodiversity in the agricultural landscape has led to interest in, and development of, environmental policies such as agri-environmental schemes.

The survey depicted scenarios of land use changes from high intensity agriculture to either low intensity agriculture or plantation of new woodlands; respondents were asked to choose one of the three options for change or the status quo. The relevant attributes of the scenarios were: 1) spatially specific attributes regarding the alternatives’ location in GB relative to the respondent’s home (i.e. distance and country in which each site was located); 2) attributes related to impact of the proposed change (i.e. impacts on biodiversity, accessibility to public, size of the sites); and 3) a price attribute capturing the cost the household would incur for each option.

The presentation of the scenarios on the screen is the novelty of the study. Each scenario was displayed on a map where respondents’ home location and the spatial locations of each option for change was made explicit.  The choice sets were tailored to respondent’s home location and used a set of sequential D-Efficient experimental designs to generate a unique map of GB for each choice situation. Actual land use change data was included in the design to offer sites where high intensity agriculture currently exists and where the proposed scenario can theoretically occur. This complex survey design setting was developed with specialized web-programmers.

The map presentation of the choice was intended to decrease the respondent’s cognitive load, by representing two attributes (country and distance) with a visual/map presentation, improving the commonly used table format. In time of a widespread use of mapping tools such as Google maps this was expected to be familiar to respondents and make the choice easier. A split sample allowed us to test the effect of map presentation on respondents’ abilities to make choices both within and across the full population, with 250 (control) subsample facing the exact same design however with the choices presented in a table format only (i.e. without maps). Initial results show that respondents are overwhelmingly in favour of map presentation and that this presentation makes the choices more realistic and more understandable for a majority of respondents. Further, the results indicate that the respondents faced with a map version of the survey were more consistent in their choices and had more strongly formed preferences, particularly for distance and country – the two spatial attributes of our design directly related to the use of maps.

Overall results show distance, size and price attributes of expected magnitudes and signs and identify that impacts on biodiversity was key to respondents’ choices. It also seems that land use policies are location sensitive and respondents attached a higher value to “home country” projects, and a prominent preference for sites being accessible to the public, reflecting presence of both use and non-use values.

Around two thirds of respondents exhibited stable preferences by choosing the same alternative for a repeated question, one which they had previously answered in the course of the survey. This is encouraging in the context of such complex choices. The survey also tested the comprehension of the main points related to the depicted scenarios and showed that the respondents exhibiting lack of comprehension of (or attention to) the survey were less consistent in their choices. Such “response quality” indicators together with an assessment of the learning and fatigue effects (using timestamps for choice questions), and a range of control questions, provide an opportunity to ensure the provision for the highest quality of evidence for policy advice.

The study also provides a contribution to the debate around and development of post-Brexit or national policy for agriculture and the environment in the UK. A number of follow-up questions in the survey relating to these topics show that a representative sample of the public (n=2500) in Great Britain wishes the overall support for environment and agriculture to increase. Importantly, a majority of respondents support a shift of overall funding towards interventions generating public goods from agriculture (such as interventions valued in this survey) rather than direct farmers’ income support.

This research contributes to land use policy advice in the UK, offers a novel approach of choice set presentation and enhances the understanding of key components of choice experiment design. In the context of an increased policy interest in the economic valuation literature, this paper is timely and represents a viable contribution to the conference proceedings. We hope that it will provide a useful point of departure for fruitful discussions on choice modelling research in environmental economics in the context of environmental valuation and how to ensure that our scientific field delivers data that is useful for national and regional policy.

Conference registration is required in order to view papers.